Where to Work in Machine Learning?
A few years ago(around 2019), I happened to stumble on a quora answer to a question similar to where to work in machine learning. The answer was something like this: “Find the place where you can solve most important problems”
She said, i.e. climate change, people who are good in understanding climate problems dont have ML expertise and people who have ML expertise are not working on climate change problems. So if you can solve climate change problems using ML, you will be in a great position.
Here, climate change is an example. You can replace it with any other problem space where ML is largely untouched. This answer was not popular on quora, but it struck a chord with me.
I believe if you want to set yourself apart in data science, you need to work on problems where ML can transform the industry. I dont know where I read it, it was like this “its sad that smartest people on earth are working on how to make people click”. So its better that we work on problems that are worth solving for.
The next question comes, How to identify such problems and places to work for? I would say keep talking to people and give interviews. Think about the problems that you want to solve. If you dont talk to a diverse set of people in different industries, you will never know what problems are worth solving. For example, I didnot know that you can use ML in energy sector to save electricity. One of the colleagues told me that he worked on an oil and gas industry project where they were using ML to predict when they should do maintenance of the machines. ML is saving lives of people on the job-site. You already know that ML is transforming healthcare industry.
The point I want to make is you need to talk to people and read more whats happening in the industry. Give interviews to the unheard(relatively unknown) companies. We dont apply to these companies because these are not popular companies to work for. But these companies might be the ones who are solving the most important problems.
Once you start exposing yourself companies other than popular companies like FAANG, you will start to realize that companies are solving important problems and can decide if you want to work there or not.
You found the industry, now what?
You know the industry where you want to work. Now you need to find the company. How to find the company?
Contrarian View
Suppose you found the company and problem space where you want to work for. Now an interesting thing happened, you got an offer from a company where you liked the people and pay is great and learning opportunity is great. But the problem space is not that interesting. What should you do? In my opinion, you can go work for this company. I would do. Recently I was asked what you want to work upon? I told them
UPDATE: I just watched andrew ng’s advice on emerging trend in AI. It largely aligns with my thoughts above. Do watch it.